feat: show top largest files

README.md

@@ -71,6 +71,7 @@ Flags:
       --si                            Show sizes with decimal SI prefixes (kB, MB, GB) instead of binary prefixes (KiB, MiB, GiB)
       --storage-path string           Path to persistent key-value storage directory (default "/tmp/badger")
   -s, --summarize                     Show only a total in non-interactive mode
+  -t, --top int                       Show only top X largest files in non-interactive mode
       --use-storage                   Use persistent key-value storage for analysis data (experimental)
   -v, --version                       Print version
       --write-config                  Write current configuration to file (default is $HOME/.gdu.yaml)
@@ -104,6 +105,7 @@ Basic list of actions in interactive mode (show help modal for more):
     gdu -n /                              # only print stats, do not start interactive mode
     gdu -np /                             # do not show progress, useful when using its output in a script
     gdu -nps /some/dir                    # show only total usage for given dir
+    gdu -nt 10 /                          # show top 10 largest files
     gdu / > file                          # write stats to file, do not start interactive mode
 
     gdu -o- / | gzip -c >report.json.gz   # write all info to JSON file for later analysis

cmd/gdu/app/app.go

@@ -72,6 +72,7 @@ type Flags struct {
 	StoragePath        string   `yaml:"storage-path"`
 	ReadFromStorage    bool     `yaml:"read-from-storage"`
 	Summarize          bool     `yaml:"summarize"`
+	Top                int      `yaml:"top"`
 	UseSIPrefix        bool     `yaml:"use-si-prefix"`
 	NoPrefix           bool     `yaml:"no-prefix"`
 	WriteConfig        bool     `yaml:"-"`
@@ -242,6 +243,7 @@ func (a *App) createUI() (UI, error) {
 			a.Flags.ConstGC,
 			a.Flags.UseSIPrefix,
 			a.Flags.NoPrefix,
+			a.Flags.Top,
 		)
 		if a.Flags.NoUnicode {
 			stdoutUI.UseOldProgressRunes()

cmd/gdu/main.go

@@ -74,6 +74,7 @@ func init() {
 	flags.BoolVarP(&af.NoProgress, "no-progress", "p", false, "Do not show progress in non-interactive mode")
 	flags.BoolVarP(&af.NoUnicode, "no-unicode", "u", false, "Do not use Unicode symbols (for size bar)")
 	flags.BoolVarP(&af.Summarize, "summarize", "s", false, "Show only a total in non-interactive mode")
+	flags.IntVarP(&af.Top, "top", "t", 0, "Show only top X largest files in non-interactive mode")
 	flags.BoolVar(&af.UseSIPrefix, "si", false, "Show sizes with decimal SI prefixes (kB, MB, GB) instead of binary prefixes (KiB, MiB, GiB)")
 	flags.BoolVar(&af.NoPrefix, "no-prefix", false, "Show sizes as raw numbers without any prefixes (SI or binary) in non-interactive mode")
 	flags.BoolVar(&af.NoMouse, "no-mouse", false, "Do not use mouse")

gdu.1.md

@@ -54,6 +54,8 @@ non-interactive mode
 
 **-s**, **\--summarize**\[=false\] Show only a total in non-interactive mode
 
+**-t**, **\--top**\[=0\] Show only top X largest files in non-interactive mode
+
 **-d**, **\--show-disks**\[=false\] Show all mounted disks
 
 **-a**, **\--show-apparent-size**\[=false\] Show apparent size

pkg/analyze/top.go

@@ -0,0 +1,53 @@
+package analyze
+
+import (
+	"sort"
+
+	"github.com/dundee/gdu/v5/pkg/fs"
+)
+
+// TopList is a list of top largest files
+type TopList struct {
+	Count   int
+	Items   fs.Files
+	MinSize int64
+}
+
+// NewTopList creates new TopList
+func NewTopList(count int) *TopList {
+	return &TopList{Count: count}
+}
+
+// Add adds file to the list
+func (tl *TopList) Add(file fs.Item) {
+	if len(tl.Items) < tl.Count {
+		tl.Items = append(tl.Items, file)
+		if file.GetSize() > tl.MinSize {
+			tl.MinSize = file.GetSize()
+		}
+	} else if file.GetSize() > tl.MinSize {
+		tl.Items = append(tl.Items, file)
+		tl.MinSize = file.GetSize()
+		if len(tl.Items) > tl.Count {
+			sort.Sort(fs.ByApparentSize(tl.Items))
+			tl.Items = tl.Items[1:]
+		}
+	}
+}
+
+func CollectTopFiles(dir fs.Item, count int) fs.Files {
+	topList := NewTopList(count)
+	walkDir(dir, topList)
+	sort.Sort(sort.Reverse(fs.ByApparentSize(topList.Items)))
+	return topList.Items
+}
+
+func walkDir(dir fs.Item, topList *TopList) {
+	for _, item := range dir.GetFiles() {
+		if item.IsDir() {
+			walkDir(item, topList)
+		} else {
+			topList.Add(item)
+		}
+	}
+}

pkg/analyze/top_test.go

@@ -0,0 +1,38 @@
+package analyze
+
+import (
+	"testing"
+
+	"github.com/dundee/gdu/v5/internal/testdir"
+	"github.com/stretchr/testify/assert"
+)
+
+func TestCollectTopFiles2(t *testing.T) {
+	fin := testdir.CreateTestDir()
+	defer fin()
+
+	dir := CreateAnalyzer().AnalyzeDir(
+		"test_dir", func(_, _ string) bool { return false }, false,
+	)
+
+	topFiles := CollectTopFiles(dir, 2)
+	assert.Equal(t, 2, len(topFiles))
+	assert.Equal(t, "file", topFiles[0].GetName())
+	assert.Equal(t, int64(5), topFiles[0].GetSize())
+	assert.Equal(t, "file2", topFiles[1].GetName())
+	assert.Equal(t, int64(2), topFiles[1].GetSize())
+}
+
+func TestCollectTopFiles1(t *testing.T) {
+	fin := testdir.CreateTestDir()
+	defer fin()
+
+	dir := CreateAnalyzer().AnalyzeDir(
+		"test_dir", func(_, _ string) bool { return false }, false,
+	)
+
+	topFiles := CollectTopFiles(dir, 1)
+	assert.Equal(t, 1, len(topFiles))
+	assert.Equal(t, "file", topFiles[0].GetName())
+	assert.Equal(t, int64(5), topFiles[0].GetSize())
+}

stdout/stdout.go

@@ -26,6 +26,7 @@
 	blue      *color.Color
 	summarize bool
 	noPrefix  bool
+	top       int
 }
 
 var (
@@ -45,6 +46,7 @@
 	constGC bool,
 	useSIPrefix bool,
 	noPrefix bool,
+	top int,
 ) *UI {
 	ui := &UI{
 		UI: &common.UI{
@@ -59,6 +61,7 @@
 		output:    output,
 		summarize: summarize,
 		noPrefix:  noPrefix,
+		top:       top,
 	}
 
 	ui.red = color.New(color.FgRed).Add(color.Bold)
@@ -167,9 +170,12 @@
 
 	wait.Wait()
 
-	if ui.summarize {
+	switch {
+	case ui.top > 0:
+		ui.printTopFiles(dir)
+	case ui.summarize:
 		ui.printTotalItem(dir)
-	} else {
+	default:
 		ui.showDir(dir)
 	}
 
@@ -187,9 +193,12 @@
 		return err
 	}
 
-	if ui.summarize {
+	switch {
+	case ui.top > 0:
+		ui.printTopFiles(dir)
+	case ui.summarize:
 		ui.printTotalItem(dir)
-	} else {
+	default:
 		ui.showDir(dir)
 	}
 	return nil
@@ -203,6 +212,13 @@
 	}
 }
 
+func (ui *UI) printTopFiles(file fs.Item) {
+	collected := analyze.CollectTopFiles(file, ui.top)
+	for _, file := range collected {
+		ui.printItemPath(file)
+	}
+}
+
 func (ui *UI) printTotalItem(file fs.Item) {
 	var lineFormat string
 	if ui.UseColors {
@@ -256,6 +272,27 @@
 	}
 }
 
+func (ui *UI) printItemPath(file fs.Item) {
+	var lineFormat string
+	if ui.UseColors {
+		lineFormat = "%20s %s\n"
+	} else {
+		lineFormat = "%9s %s\n"
+	}
+
+	var size int64
+	if ui.ShowApparentSize {
+		size = file.GetSize()
+	} else {
+		size = file.GetUsage()
+	}
+
+	fmt.Fprintf(ui.output,
+		lineFormat,
+		ui.formatSize(size),
+		file.GetPath())
+}
+
 // ReadAnalysis reads analysis report from JSON file
 func (ui *UI) ReadAnalysis(input io.Reader) error {
 	var (

stdout/stdout_linux_test.go

@@ -21,7 +21,7 @@ func TestShowDevicesWithErr(t *testing.T) {
 	output := bytes.NewBuffer(make([]byte, 10))
 
 	getter := device.LinuxDevicesInfoGetter{MountsPath: "/xyzxyz"}
-	ui := CreateStdoutUI(output, false, true, false, false, false, false, false, false)
+	ui := CreateStdoutUI(output, false, true, false, false, false, false, false, false, 0)
 	err := ui.ListDevices(getter)
 
 	assert.Contains(t, err.Error(), "no such file")